Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
#data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Found mnist Data
Found celeba Data

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7f354a90f978>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7f354a82f748>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.1.0
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    inputs_real = tf.placeholder(tf.float32, (None, image_width, image_height, image_channels), name='input_real')
    inputs_z = tf.placeholder(tf.float32, (None, z_dim), name='input_z')
    learning_rate = tf.placeholder(tf.float32, name='learning_rate')

    return inputs_real, inputs_z, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

def discriminator(images, reuse=False, alpha=0.2):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    with tf.variable_scope('discriminator', reuse=reuse):
        
        # 1st convolutional layer.
        x1 = tf.layers.conv2d(images, filters=64, kernel_size=5, strides=(2,2), padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Batch normalization. DON'T USE HERE, CREATES ARTEFACTS
         # Leaky ReLU.
        relu1 = tf.maximum(alpha * x1, x1)
        
        # 2nd convolutional layer.
        x2 = tf.layers.conv2d(relu1, filters=128, kernel_size=5, strides=(2,2), padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Batch normalization.
        b2 = tf.layers.batch_normalization(x2, training=True)
         # Leaky ReLU.
        relu2 = tf.maximum(alpha * b2, b2)
        
        # 3rd convolutional layer.
        x3 = tf.layers.conv2d(relu2, filters=256, kernel_size=5, strides=(2,2), padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Batch normalization.
        b3 = tf.layers.batch_normalization(x3, training=True)
         # Leaky ReLU.
        relu3 = tf.maximum(alpha * b3, b3)
        
        # Turn last layer into vector.
        x4 = tf.reshape(relu3, (-1, 4*4*256))
        logits = tf.layers.dense(x4, 1, kernel_initializer=tf.contrib.layers.xavier_initializer())
        out = tf.sigmoid(logits)
        
        return out, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, training=True, alpha=0.2):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    with tf.variable_scope('generator', reuse=not training):
        # First fully connected layer
        # TF does not have leaky ReLU built in, so we use no
        # activation fuction but create it afterwards
        h1 = tf.layers.dense(z, 4*4*512, activation=None, kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Reshape into 3d array to start the convolutional part of the network.
        h1 = tf.reshape(h1, (-1, 4, 4, 512))
        # Batch normalization.
        h1 = tf.layers.batch_normalization(h1, training=training)
        # Leaky ReLU activation function.
        h1 = tf.maximum(alpha * h1, h1)
        # h1 now has a dimension of 4x4x512
        
        # 2nd convolutional layer 7x7x256
        h2 = tf.layers.conv2d_transpose(h1, filters=256, kernel_size=4, strides=(1,1), padding='valid', kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Batch normalization.
        h2 = tf.layers.batch_normalization(h2, training=training)
         # Leaky ReLU.
        h2 = tf.maximum(alpha * h2, h2)
         
        # 3rd convolutional layer 14x14x128
        h3 = tf.layers.conv2d_transpose(h2, filters=128, kernel_size=5, strides=(2,2), padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
        # Batch normalization.
        h3 = tf.layers.batch_normalization(h3, training=training)
         # Leaky ReLU.
        h3 = tf.maximum(alpha * h3, h3)
        
        # Output layer, 28x28x3
        logits = tf.layers.conv2d_transpose(inputs=h3, filters=out_channel_dim, kernel_size=5, strides=(2,2), padding='same', kernel_initializer=tf.contrib.layers.xavier_initializer())
        out = tf.tanh(logits)
        
        return out


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim, alpha=0.2):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    g_model = generator(input_z, out_channel_dim, alpha=alpha)
    d_model_real, d_logits_real = discriminator(input_real, alpha=alpha)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse=True, alpha=alpha)
    
    smooth = 0.1
    d_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_real, labels = tf.ones_like((d_model_real) * (1 - smooth))))#labels=tf.ones_like(d_model_real)))
    d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=d_logits_fake, labels=tf.ones_like(d_model_fake)))

    d_loss = d_loss_real + d_loss_fake

    return d_loss, g_loss
    


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    # Get weights and bias to update
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith('discriminator')]
    g_vars = [var for var in t_vars if var.name.startswith('generator')]

    # Optimize
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(d_loss, var_list=d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1=beta1).minimize(g_loss, var_list=g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np
import matplotlib
def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode, losses):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    
    matplotlib.rcParams['figure.figsize'] = [10,6]
    pyplot.subplot(1, 2, 1)
    pyplot.plot(range(len(losses)), [i[1] for i in losses], '-g', label='Gen loss')
    pyplot.plot(range(len(losses)), [i[0] for i in losses], '-r', label='Dis loss')
    pyplot.xlabel("Step")
    pyplot.ylabel("Loss")
    pyplot.legend()
    pyplot.grid()
    pyplot.gca().set_xlim((0,len(losses)))
    pyplot.gca().set_ylim((0,10))
    pyplot.gca().set_aspect(len(losses)/10.)
    pyplot.subplot(1, 2, 2)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

import time
from matplotlib import pyplot as plt
def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    # Build model.
    # Create inputs.
    input_real, input_z, learning_rate_placeholder = model_inputs(data_shape[1], data_shape[2], data_shape[3], z_dim)
    # Create losses.
    d_loss, g_loss = model_loss(input_real, input_z, data_shape[3], alpha=0.2)
    # Create optimizers.
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate_placeholder, beta1)

    print_every = 10
    show_every = 100
        
    losses = []
    steps = 0
        
    time_start = time.time()
    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                steps += 1
                # Rescale batch images to range from -1 to 1
                batch_images = batch_images * 2
                
                # Sample random noise for G
                batch_z = np.random.uniform(-1, 1, size=(batch_size, z_dim))

                # Run optimizers
                _ = sess.run(d_opt, feed_dict={input_real: batch_images, input_z: batch_z, learning_rate_placeholder:learning_rate})
                _ = sess.run(g_opt, feed_dict={input_z: batch_z, input_real: batch_images, learning_rate_placeholder:learning_rate})

                if steps % print_every == 0:
                    # At the end of each epoch, get the losses and print them out
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})

                    time_elapsed = time.time() - time_start
                    progress = steps / (epoch_count * (data_shape[0]/batch_size))
                    time_to_finish = time_elapsed / progress - time_elapsed
                    print("Epc: {}/{}    ".format(epoch_i+1, epochs) +
                          "Bat: {:04g}/{}    ".format(steps - epoch_i*int(data_shape[0]/batch_size), int(data_shape[0]/batch_size)) +
                          "Dis Loss: {:.4f}    ".format(train_loss_d) +
                          "Gen Loss: {:.4f}    ".format(train_loss_g) +
                          "Time elaps: {}    ".format(time.strftime("%H:%M:%S", time.gmtime(time_elapsed))) +
                          "Time remain: {}.".format(time.strftime("%H:%M:%S", time.gmtime(time_to_finish))))
                    # Save losses to view after training
                    losses.append((train_loss_d, train_loss_g))

                if steps % show_every == 0:
                    show_generator_output(sess, 25, input_z, data_shape[3], data_image_mode, losses)
                    '''
                    plt.plot(range(len(losses)), [i[0] for i in losses], '-r', label='Dis loss')
                    plt.plot(range(len(losses)), [i[1] for i in losses], '-g', label='Gen loss')
                    plt.xlabel("Step")
                    plt.ylabel("Loss")
                    plt.legend()
                    plt.show()
                    '''

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 50
z_dim = 100
learning_rate = 0.001
beta1 = 0.5


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches, mnist_dataset.shape, mnist_dataset.image_mode)

Epc: 1/2    Bat: 0010/1200    Dis Loss: 1.3803    Gen Loss: 15.6087    Time elaps: 00:00:01    Time remain: 00:07:41.
Epc: 1/2    Bat: 0020/1200    Dis Loss: 0.0921    Gen Loss: 16.8544    Time elaps: 00:00:03    Time remain: 00:06:45.
Epc: 1/2    Bat: 0030/1200    Dis Loss: 0.1494    Gen Loss: 3.6232    Time elaps: 00:00:04    Time remain: 00:06:26.
Epc: 1/2    Bat: 0040/1200    Dis Loss: 1.2054    Gen Loss: 0.8663    Time elaps: 00:00:06    Time remain: 00:06:16.
Epc: 1/2    Bat: 0050/1200    Dis Loss: 0.3780    Gen Loss: 3.3904    Time elaps: 00:00:07    Time remain: 00:06:10.
Epc: 1/2    Bat: 0060/1200    Dis Loss: 0.9392    Gen Loss: 5.5800    Time elaps: 00:00:09    Time remain: 00:06:05.
Epc: 1/2    Bat: 0070/1200    Dis Loss: 0.8240    Gen Loss: 1.2972    Time elaps: 00:00:10    Time remain: 00:06:01.
Epc: 1/2    Bat: 0080/1200    Dis Loss: 0.7434    Gen Loss: 1.1407    Time elaps: 00:00:12    Time remain: 00:05:57.
Epc: 1/2    Bat: 0090/1200    Dis Loss: 0.4804    Gen Loss: 3.0844    Time elaps: 00:00:13    Time remain: 00:05:54.
Epc: 1/2    Bat: 0100/1200    Dis Loss: 0.2901    Gen Loss: 3.5318    Time elaps: 00:00:15    Time remain: 00:05:52.

Epc: 1/2    Bat: 0110/1200    Dis Loss: 3.2211    Gen Loss: 0.0988    Time elaps: 00:00:17    Time remain: 00:05:56.
Epc: 1/2    Bat: 0120/1200    Dis Loss: 1.7288    Gen Loss: 0.4792    Time elaps: 00:00:18    Time remain: 00:05:53.
Epc: 1/2    Bat: 0130/1200    Dis Loss: 0.8124    Gen Loss: 1.0478    Time elaps: 00:00:20    Time remain: 00:05:51.
Epc: 1/2    Bat: 0140/1200    Dis Loss: 0.7756    Gen Loss: 1.0994    Time elaps: 00:00:21    Time remain: 00:05:48.
Epc: 1/2    Bat: 0150/1200    Dis Loss: 1.7023    Gen Loss: 3.2023    Time elaps: 00:00:23    Time remain: 00:05:46.
Epc: 1/2    Bat: 0160/1200    Dis Loss: 1.1001    Gen Loss: 0.9400    Time elaps: 00:00:24    Time remain: 00:05:43.
Epc: 1/2    Bat: 0170/1200    Dis Loss: 0.6843    Gen Loss: 2.4020    Time elaps: 00:00:26    Time remain: 00:05:41.
Epc: 1/2    Bat: 0180/1200    Dis Loss: 0.7727    Gen Loss: 1.5815    Time elaps: 00:00:27    Time remain: 00:05:39.
Epc: 1/2    Bat: 0190/1200    Dis Loss: 0.7709    Gen Loss: 1.4576    Time elaps: 00:00:29    Time remain: 00:05:37.
Epc: 1/2    Bat: 0200/1200    Dis Loss: 0.7219    Gen Loss: 1.1451    Time elaps: 00:00:30    Time remain: 00:05:35.

Epc: 1/2    Bat: 0210/1200    Dis Loss: 1.3769    Gen Loss: 0.5371    Time elaps: 00:00:32    Time remain: 00:05:37.
Epc: 1/2    Bat: 0220/1200    Dis Loss: 0.9298    Gen Loss: 1.1109    Time elaps: 00:00:33    Time remain: 00:05:35.
Epc: 1/2    Bat: 0230/1200    Dis Loss: 1.1909    Gen Loss: 0.5806    Time elaps: 00:00:35    Time remain: 00:05:33.
Epc: 1/2    Bat: 0240/1200    Dis Loss: 1.2984    Gen Loss: 0.6414    Time elaps: 00:00:36    Time remain: 00:05:31.
Epc: 1/2    Bat: 0250/1200    Dis Loss: 0.6997    Gen Loss: 1.7400    Time elaps: 00:00:38    Time remain: 00:05:29.
Epc: 1/2    Bat: 0260/1200    Dis Loss: 0.7748    Gen Loss: 1.5103    Time elaps: 00:00:39    Time remain: 00:05:27.
Epc: 1/2    Bat: 0270/1200    Dis Loss: 1.0903    Gen Loss: 1.1198    Time elaps: 00:00:41    Time remain: 00:05:25.
Epc: 1/2    Bat: 0280/1200    Dis Loss: 0.7907    Gen Loss: 1.7673    Time elaps: 00:00:42    Time remain: 00:05:23.
Epc: 1/2    Bat: 0290/1200    Dis Loss: 1.2899    Gen Loss: 0.5885    Time elaps: 00:00:44    Time remain: 00:05:21.
Epc: 1/2    Bat: 0300/1200    Dis Loss: 0.9317    Gen Loss: 1.4770    Time elaps: 00:00:45    Time remain: 00:05:19.

Epc: 1/2    Bat: 0310/1200    Dis Loss: 1.1012    Gen Loss: 0.9444    Time elaps: 00:00:47    Time remain: 00:05:20.
Epc: 1/2    Bat: 0320/1200    Dis Loss: 1.5997    Gen Loss: 2.4166    Time elaps: 00:00:48    Time remain: 00:05:18.
Epc: 1/2    Bat: 0330/1200    Dis Loss: 1.0090    Gen Loss: 0.8308    Time elaps: 00:00:50    Time remain: 00:05:16.
Epc: 1/2    Bat: 0340/1200    Dis Loss: 1.6913    Gen Loss: 0.3545    Time elaps: 00:00:51    Time remain: 00:05:14.
Epc: 1/2    Bat: 0350/1200    Dis Loss: 1.1453    Gen Loss: 1.5473    Time elaps: 00:00:53    Time remain: 00:05:12.
Epc: 1/2    Bat: 0360/1200    Dis Loss: 0.7182    Gen Loss: 1.4445    Time elaps: 00:00:54    Time remain: 00:05:11.
Epc: 1/2    Bat: 0370/1200    Dis Loss: 1.1259    Gen Loss: 0.7035    Time elaps: 00:00:56    Time remain: 00:05:09.
Epc: 1/2    Bat: 0380/1200    Dis Loss: 0.8969    Gen Loss: 1.3961    Time elaps: 00:00:57    Time remain: 00:05:07.
Epc: 1/2    Bat: 0390/1200    Dis Loss: 0.9621    Gen Loss: 1.7501    Time elaps: 00:00:59    Time remain: 00:05:05.
Epc: 1/2    Bat: 0400/1200    Dis Loss: 1.3405    Gen Loss: 1.4459    Time elaps: 00:01:00    Time remain: 00:05:04.

Epc: 1/2    Bat: 0410/1200    Dis Loss: 1.0069    Gen Loss: 0.8511    Time elaps: 00:01:02    Time remain: 00:05:04.
Epc: 1/2    Bat: 0420/1200    Dis Loss: 0.9579    Gen Loss: 0.8740    Time elaps: 00:01:04    Time remain: 00:05:02.
Epc: 1/2    Bat: 0430/1200    Dis Loss: 1.7757    Gen Loss: 0.3104    Time elaps: 00:01:05    Time remain: 00:05:00.
Epc: 1/2    Bat: 0440/1200    Dis Loss: 1.4670    Gen Loss: 0.4579    Time elaps: 00:01:07    Time remain: 00:04:59.
Epc: 1/2    Bat: 0450/1200    Dis Loss: 0.9011    Gen Loss: 1.5641    Time elaps: 00:01:08    Time remain: 00:04:57.
Epc: 1/2    Bat: 0460/1200    Dis Loss: 1.3546    Gen Loss: 1.7307    Time elaps: 00:01:10    Time remain: 00:04:55.
Epc: 1/2    Bat: 0470/1200    Dis Loss: 1.3059    Gen Loss: 2.0295    Time elaps: 00:01:11    Time remain: 00:04:53.
Epc: 1/2    Bat: 0480/1200    Dis Loss: 1.0204    Gen Loss: 0.9073    Time elaps: 00:01:13    Time remain: 00:04:52.
Epc: 1/2    Bat: 0490/1200    Dis Loss: 1.2180    Gen Loss: 0.5205    Time elaps: 00:01:14    Time remain: 00:04:50.
Epc: 1/2    Bat: 0500/1200    Dis Loss: 1.1065    Gen Loss: 0.8990    Time elaps: 00:01:16    Time remain: 00:04:48.

Epc: 1/2    Bat: 0510/1200    Dis Loss: 0.9068    Gen Loss: 1.3164    Time elaps: 00:01:17    Time remain: 00:04:48.
Epc: 1/2    Bat: 0520/1200    Dis Loss: 1.3871    Gen Loss: 0.4686    Time elaps: 00:01:19    Time remain: 00:04:46.
Epc: 1/2    Bat: 0530/1200    Dis Loss: 1.0580    Gen Loss: 0.6627    Time elaps: 00:01:20    Time remain: 00:04:45.
Epc: 1/2    Bat: 0540/1200    Dis Loss: 0.9612    Gen Loss: 1.0237    Time elaps: 00:01:22    Time remain: 00:04:43.
Epc: 1/2    Bat: 0550/1200    Dis Loss: 1.2233    Gen Loss: 1.8385    Time elaps: 00:01:23    Time remain: 00:04:41.
Epc: 1/2    Bat: 0560/1200    Dis Loss: 1.0076    Gen Loss: 0.9470    Time elaps: 00:01:25    Time remain: 00:04:40.
Epc: 1/2    Bat: 0570/1200    Dis Loss: 1.0595    Gen Loss: 0.8355    Time elaps: 00:01:26    Time remain: 00:04:38.
Epc: 1/2    Bat: 0580/1200    Dis Loss: 1.0752    Gen Loss: 1.0000    Time elaps: 00:01:28    Time remain: 00:04:36.
Epc: 1/2    Bat: 0590/1200    Dis Loss: 1.0151    Gen Loss: 1.1892    Time elaps: 00:01:29    Time remain: 00:04:35.
Epc: 1/2    Bat: 0600/1200    Dis Loss: 1.1949    Gen Loss: 0.5443    Time elaps: 00:01:31    Time remain: 00:04:33.

Epc: 1/2    Bat: 0610/1200    Dis Loss: 1.1841    Gen Loss: 0.7328    Time elaps: 00:01:33    Time remain: 00:04:33.
Epc: 1/2    Bat: 0620/1200    Dis Loss: 1.1045    Gen Loss: 0.6590    Time elaps: 00:01:34    Time remain: 00:04:31.
Epc: 1/2    Bat: 0630/1200    Dis Loss: 1.4282    Gen Loss: 0.4115    Time elaps: 00:01:36    Time remain: 00:04:29.
Epc: 1/2    Bat: 0640/1200    Dis Loss: 1.5644    Gen Loss: 1.7885    Time elaps: 00:01:37    Time remain: 00:04:28.
Epc: 1/2    Bat: 0650/1200    Dis Loss: 1.0385    Gen Loss: 1.2066    Time elaps: 00:01:39    Time remain: 00:04:26.
Epc: 1/2    Bat: 0660/1200    Dis Loss: 1.3861    Gen Loss: 0.4145    Time elaps: 00:01:40    Time remain: 00:04:24.
Epc: 1/2    Bat: 0670/1200    Dis Loss: 0.9561    Gen Loss: 1.0052    Time elaps: 00:01:41    Time remain: 00:04:23.
Epc: 1/2    Bat: 0680/1200    Dis Loss: 1.0579    Gen Loss: 0.6885    Time elaps: 00:01:43    Time remain: 00:04:21.
Epc: 1/2    Bat: 0690/1200    Dis Loss: 0.8568    Gen Loss: 1.2625    Time elaps: 00:01:44    Time remain: 00:04:20.
Epc: 1/2    Bat: 0700/1200    Dis Loss: 1.0199    Gen Loss: 0.6841    Time elaps: 00:01:46    Time remain: 00:04:18.

Epc: 1/2    Bat: 0710/1200    Dis Loss: 1.0448    Gen Loss: 0.7892    Time elaps: 00:01:48    Time remain: 00:04:17.
Epc: 1/2    Bat: 0720/1200    Dis Loss: 1.0871    Gen Loss: 1.1298    Time elaps: 00:01:49    Time remain: 00:04:16.
Epc: 1/2    Bat: 0730/1200    Dis Loss: 1.1108    Gen Loss: 0.7027    Time elaps: 00:01:51    Time remain: 00:04:14.
Epc: 1/2    Bat: 0740/1200    Dis Loss: 1.0750    Gen Loss: 0.5736    Time elaps: 00:01:52    Time remain: 00:04:12.
Epc: 1/2    Bat: 0750/1200    Dis Loss: 1.1300    Gen Loss: 1.4787    Time elaps: 00:01:54    Time remain: 00:04:11.
Epc: 1/2    Bat: 0760/1200    Dis Loss: 0.9171    Gen Loss: 1.2783    Time elaps: 00:01:55    Time remain: 00:04:09.
Epc: 1/2    Bat: 0770/1200    Dis Loss: 1.2474    Gen Loss: 1.5981    Time elaps: 00:01:57    Time remain: 00:04:08.
Epc: 1/2    Bat: 0780/1200    Dis Loss: 1.1812    Gen Loss: 0.5762    Time elaps: 00:01:58    Time remain: 00:04:06.
Epc: 1/2    Bat: 0790/1200    Dis Loss: 0.8654    Gen Loss: 1.3199    Time elaps: 00:02:00    Time remain: 00:04:04.
Epc: 1/2    Bat: 0800/1200    Dis Loss: 1.2456    Gen Loss: 1.1514    Time elaps: 00:02:01    Time remain: 00:04:03.

Epc: 1/2    Bat: 0810/1200    Dis Loss: 1.0123    Gen Loss: 1.1352    Time elaps: 00:02:03    Time remain: 00:04:02.
Epc: 1/2    Bat: 0820/1200    Dis Loss: 1.3812    Gen Loss: 0.4420    Time elaps: 00:02:04    Time remain: 00:04:00.
Epc: 1/2    Bat: 0830/1200    Dis Loss: 1.6102    Gen Loss: 0.2764    Time elaps: 00:02:06    Time remain: 00:03:59.
Epc: 1/2    Bat: 0840/1200    Dis Loss: 0.9812    Gen Loss: 0.7228    Time elaps: 00:02:07    Time remain: 00:03:57.
Epc: 1/2    Bat: 0850/1200    Dis Loss: 1.4186    Gen Loss: 0.5691    Time elaps: 00:02:09    Time remain: 00:03:55.
Epc: 1/2    Bat: 0860/1200    Dis Loss: 1.2501    Gen Loss: 0.5674    Time elaps: 00:02:10    Time remain: 00:03:54.
Epc: 1/2    Bat: 0870/1200    Dis Loss: 0.9499    Gen Loss: 1.0110    Time elaps: 00:02:12    Time remain: 00:03:52.
Epc: 1/2    Bat: 0880/1200    Dis Loss: 1.0801    Gen Loss: 0.6573    Time elaps: 00:02:13    Time remain: 00:03:51.
Epc: 1/2    Bat: 0890/1200    Dis Loss: 1.0220    Gen Loss: 0.7285    Time elaps: 00:02:15    Time remain: 00:03:49.
Epc: 1/2    Bat: 0900/1200    Dis Loss: 1.1407    Gen Loss: 0.5721    Time elaps: 00:02:16    Time remain: 00:03:48.

Epc: 1/2    Bat: 0910/1200    Dis Loss: 1.1611    Gen Loss: 0.5487    Time elaps: 00:02:18    Time remain: 00:03:46.
Epc: 1/2    Bat: 0920/1200    Dis Loss: 0.9817    Gen Loss: 1.0147    Time elaps: 00:02:20    Time remain: 00:03:45.
Epc: 1/2    Bat: 0930/1200    Dis Loss: 1.0317    Gen Loss: 0.7400    Time elaps: 00:02:21    Time remain: 00:03:43.
Epc: 1/2    Bat: 0940/1200    Dis Loss: 0.9518    Gen Loss: 1.6158    Time elaps: 00:02:23    Time remain: 00:03:42.
Epc: 1/2    Bat: 0950/1200    Dis Loss: 1.5487    Gen Loss: 0.3332    Time elaps: 00:02:24    Time remain: 00:03:40.
Epc: 1/2    Bat: 0960/1200    Dis Loss: 1.0107    Gen Loss: 0.8481    Time elaps: 00:02:26    Time remain: 00:03:39.
Epc: 1/2    Bat: 0970/1200    Dis Loss: 0.9225    Gen Loss: 0.8400    Time elaps: 00:02:27    Time remain: 00:03:37.
Epc: 1/2    Bat: 0980/1200    Dis Loss: 1.3322    Gen Loss: 1.6891    Time elaps: 00:02:28    Time remain: 00:03:35.
Epc: 1/2    Bat: 0990/1200    Dis Loss: 1.0515    Gen Loss: 1.1116    Time elaps: 00:02:30    Time remain: 00:03:34.
Epc: 1/2    Bat: 1000/1200    Dis Loss: 1.2439    Gen Loss: 0.5045    Time elaps: 00:02:31    Time remain: 00:03:32.

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size =16
z_dim = 100
learning_rate = 0.0001
beta1 = 0.3


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 1

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Epc: 1/1    Bat: 0010/12662    Dis Loss: 2.1546    Gen Loss: 0.1606    Time elaps: 00:00:01    Time remain: 00:24:44.
Epc: 1/1    Bat: 0020/12662    Dis Loss: 1.3868    Gen Loss: 0.4053    Time elaps: 00:00:01    Time remain: 00:20:12.
Epc: 1/1    Bat: 0030/12662    Dis Loss: 1.1835    Gen Loss: 0.4550    Time elaps: 00:00:02    Time remain: 00:18:41.
Epc: 1/1    Bat: 0040/12662    Dis Loss: 0.7408    Gen Loss: 0.8861    Time elaps: 00:00:03    Time remain: 00:17:54.
Epc: 1/1    Bat: 0050/12662    Dis Loss: 1.0579    Gen Loss: 0.5992    Time elaps: 00:00:04    Time remain: 00:17:26.
Epc: 1/1    Bat: 0060/12662    Dis Loss: 0.4142    Gen Loss: 1.3272    Time elaps: 00:00:04    Time remain: 00:17:07.
Epc: 1/1    Bat: 0070/12662    Dis Loss: 0.6332    Gen Loss: 0.9612    Time elaps: 00:00:05    Time remain: 00:16:53.
Epc: 1/1    Bat: 0080/12662    Dis Loss: 0.2636    Gen Loss: 2.3322    Time elaps: 00:00:06    Time remain: 00:16:43.
Epc: 1/1    Bat: 0090/12662    Dis Loss: 0.3658    Gen Loss: 1.4207    Time elaps: 00:00:07    Time remain: 00:16:35.
Epc: 1/1    Bat: 0100/12662    Dis Loss: 1.5688    Gen Loss: 0.3006    Time elaps: 00:00:07    Time remain: 00:16:28.

Epc: 1/1    Bat: 0110/12662    Dis Loss: 0.4216    Gen Loss: 2.1093    Time elaps: 00:00:08    Time remain: 00:17:02.
Epc: 1/1    Bat: 0120/12662    Dis Loss: 0.5186    Gen Loss: 1.2191    Time elaps: 00:00:09    Time remain: 00:16:58.
Epc: 1/1    Bat: 0130/12662    Dis Loss: 0.2779    Gen Loss: 1.8991    Time elaps: 00:00:10    Time remain: 00:16:50.
Epc: 1/1    Bat: 0140/12662    Dis Loss: 1.2103    Gen Loss: 0.4710    Time elaps: 00:00:11    Time remain: 00:16:44.
Epc: 1/1    Bat: 0150/12662    Dis Loss: 0.3921    Gen Loss: 1.4236    Time elaps: 00:00:11    Time remain: 00:16:38.
Epc: 1/1    Bat: 0160/12662    Dis Loss: 0.5259    Gen Loss: 0.9594    Time elaps: 00:00:12    Time remain: 00:16:33.
Epc: 1/1    Bat: 0170/12662    Dis Loss: 0.1428    Gen Loss: 4.4037    Time elaps: 00:00:13    Time remain: 00:16:29.
Epc: 1/1    Bat: 0180/12662    Dis Loss: 0.2458    Gen Loss: 2.1533    Time elaps: 00:00:14    Time remain: 00:16:24.
Epc: 1/1    Bat: 0190/12662    Dis Loss: 0.5558    Gen Loss: 1.1525    Time elaps: 00:00:14    Time remain: 00:16:21.
Epc: 1/1    Bat: 0200/12662    Dis Loss: 0.1736    Gen Loss: 2.2846    Time elaps: 00:00:15    Time remain: 00:16:17.

Epc: 1/1    Bat: 0210/12662    Dis Loss: 0.1526    Gen Loss: 6.4571    Time elaps: 00:00:16    Time remain: 00:16:35.
Epc: 1/1    Bat: 0220/12662    Dis Loss: 0.1468    Gen Loss: 2.3470    Time elaps: 00:00:17    Time remain: 00:16:31.
Epc: 1/1    Bat: 0230/12662    Dis Loss: 0.1371    Gen Loss: 2.4331    Time elaps: 00:00:18    Time remain: 00:16:30.
Epc: 1/1    Bat: 0240/12662    Dis Loss: 0.0904    Gen Loss: 3.0466    Time elaps: 00:00:19    Time remain: 00:16:30.
Epc: 1/1    Bat: 0250/12662    Dis Loss: 0.0357    Gen Loss: 5.8969    Time elaps: 00:00:19    Time remain: 00:16:29.
Epc: 1/1    Bat: 0260/12662    Dis Loss: 0.1863    Gen Loss: 2.1726    Time elaps: 00:00:20    Time remain: 00:16:28.
Epc: 1/1    Bat: 0270/12662    Dis Loss: 0.0854    Gen Loss: 2.7004    Time elaps: 00:00:21    Time remain: 00:16:26.
Epc: 1/1    Bat: 0280/12662    Dis Loss: 0.2629    Gen Loss: 1.7988    Time elaps: 00:00:22    Time remain: 00:16:25.
Epc: 1/1    Bat: 0290/12662    Dis Loss: 0.1265    Gen Loss: 2.3658    Time elaps: 00:00:23    Time remain: 00:16:23.
Epc: 1/1    Bat: 0300/12662    Dis Loss: 0.0973    Gen Loss: 2.7502    Time elaps: 00:00:23    Time remain: 00:16:21.

Epc: 1/1    Bat: 0310/12662    Dis Loss: 0.0735    Gen Loss: 2.8462    Time elaps: 00:00:25    Time remain: 00:16:52.
Epc: 1/1    Bat: 0320/12662    Dis Loss: 0.2480    Gen Loss: 1.9818    Time elaps: 00:00:26    Time remain: 00:16:51.
Epc: 1/1    Bat: 0330/12662    Dis Loss: 0.0956    Gen Loss: 2.5629    Time elaps: 00:00:27    Time remain: 00:16:49.
Epc: 1/1    Bat: 0340/12662    Dis Loss: 0.1616    Gen Loss: 2.2670    Time elaps: 00:00:27    Time remain: 00:16:46.
Epc: 1/1    Bat: 0350/12662    Dis Loss: 0.2892    Gen Loss: 1.5975    Time elaps: 00:00:28    Time remain: 00:16:42.
Epc: 1/1    Bat: 0360/12662    Dis Loss: 0.0522    Gen Loss: 3.4598    Time elaps: 00:00:29    Time remain: 00:16:39.
Epc: 1/1    Bat: 0370/12662    Dis Loss: 0.0932    Gen Loss: 2.6281    Time elaps: 00:00:29    Time remain: 00:16:36.
Epc: 1/1    Bat: 0380/12662    Dis Loss: 0.1299    Gen Loss: 2.5362    Time elaps: 00:00:30    Time remain: 00:16:33.
Epc: 1/1    Bat: 0390/12662    Dis Loss: 0.2752    Gen Loss: 1.6818    Time elaps: 00:00:31    Time remain: 00:16:30.
Epc: 1/1    Bat: 0400/12662    Dis Loss: 0.3825    Gen Loss: 8.8131    Time elaps: 00:00:32    Time remain: 00:16:28.

Epc: 1/1    Bat: 0410/12662    Dis Loss: 0.0897    Gen Loss: 3.6229    Time elaps: 00:00:33    Time remain: 00:16:37.
Epc: 1/1    Bat: 0420/12662    Dis Loss: 0.1720    Gen Loss: 3.0897    Time elaps: 00:00:34    Time remain: 00:16:34.
Epc: 1/1    Bat: 0430/12662    Dis Loss: 0.1004    Gen Loss: 2.7676    Time elaps: 00:00:34    Time remain: 00:16:32.
Epc: 1/1    Bat: 0440/12662    Dis Loss: 0.2027    Gen Loss: 2.3036    Time elaps: 00:00:35    Time remain: 00:16:29.
Epc: 1/1    Bat: 0450/12662    Dis Loss: 0.3353    Gen Loss: 3.7400    Time elaps: 00:00:36    Time remain: 00:16:26.
Epc: 1/1    Bat: 0460/12662    Dis Loss: 0.6587    Gen Loss: 2.3974    Time elaps: 00:00:37    Time remain: 00:16:24.
Epc: 1/1    Bat: 0470/12662    Dis Loss: 0.5809    Gen Loss: 1.1181    Time elaps: 00:00:37    Time remain: 00:16:21.
Epc: 1/1    Bat: 0480/12662    Dis Loss: 0.6132    Gen Loss: 3.3115    Time elaps: 00:00:38    Time remain: 00:16:19.
Epc: 1/1    Bat: 0490/12662    Dis Loss: 0.3953    Gen Loss: 1.6691    Time elaps: 00:00:39    Time remain: 00:16:18.
Epc: 1/1    Bat: 0500/12662    Dis Loss: 0.6549    Gen Loss: 1.2188    Time elaps: 00:00:40    Time remain: 00:16:16.

Epc: 1/1    Bat: 0510/12662    Dis Loss: 0.2371    Gen Loss: 2.5616    Time elaps: 00:00:41    Time remain: 00:16:24.
Epc: 1/1    Bat: 0520/12662    Dis Loss: 0.3405    Gen Loss: 2.0388    Time elaps: 00:00:42    Time remain: 00:16:22.
Epc: 1/1    Bat: 0530/12662    Dis Loss: 0.8946    Gen Loss: 0.7030    Time elaps: 00:00:42    Time remain: 00:16:19.
Epc: 1/1    Bat: 0540/12662    Dis Loss: 0.4795    Gen Loss: 1.5679    Time elaps: 00:00:43    Time remain: 00:16:17.
Epc: 1/1    Bat: 0550/12662    Dis Loss: 0.4897    Gen Loss: 1.3262    Time elaps: 00:00:44    Time remain: 00:16:15.
Epc: 1/1    Bat: 0560/12662    Dis Loss: 0.6793    Gen Loss: 0.8643    Time elaps: 00:00:45    Time remain: 00:16:13.
Epc: 1/1    Bat: 0570/12662    Dis Loss: 1.3143    Gen Loss: 0.4379    Time elaps: 00:00:45    Time remain: 00:16:11.
Epc: 1/1    Bat: 0580/12662    Dis Loss: 0.4078    Gen Loss: 1.5258    Time elaps: 00:00:46    Time remain: 00:16:09.
Epc: 1/1    Bat: 0590/12662    Dis Loss: 0.5659    Gen Loss: 1.8762    Time elaps: 00:00:47    Time remain: 00:16:07.
Epc: 1/1    Bat: 0600/12662    Dis Loss: 0.4476    Gen Loss: 2.1149    Time elaps: 00:00:48    Time remain: 00:16:05.

Epc: 1/1    Bat: 0610/12662    Dis Loss: 0.6560    Gen Loss: 1.4097    Time elaps: 00:00:49    Time remain: 00:16:11.
Epc: 1/1    Bat: 0620/12662    Dis Loss: 0.9872    Gen Loss: 0.7338    Time elaps: 00:00:49    Time remain: 00:16:09.
Epc: 1/1    Bat: 0630/12662    Dis Loss: 1.1822    Gen Loss: 0.5752    Time elaps: 00:00:50    Time remain: 00:16:07.
Epc: 1/1    Bat: 0640/12662    Dis Loss: 0.9398    Gen Loss: 0.9376    Time elaps: 00:00:51    Time remain: 00:16:05.
Epc: 1/1    Bat: 0650/12662    Dis Loss: 1.4211    Gen Loss: 4.2911    Time elaps: 00:00:52    Time remain: 00:16:03.
Epc: 1/1    Bat: 0660/12662    Dis Loss: 0.9062    Gen Loss: 0.7478    Time elaps: 00:00:52    Time remain: 00:16:01.
Epc: 1/1    Bat: 0670/12662    Dis Loss: 1.6487    Gen Loss: 0.2784    Time elaps: 00:00:53    Time remain: 00:15:59.
Epc: 1/1    Bat: 0680/12662    Dis Loss: 0.7394    Gen Loss: 1.5794    Time elaps: 00:00:54    Time remain: 00:15:57.
Epc: 1/1    Bat: 0690/12662    Dis Loss: 0.6966    Gen Loss: 1.0521    Time elaps: 00:00:55    Time remain: 00:15:56.
Epc: 1/1    Bat: 0700/12662    Dis Loss: 1.1466    Gen Loss: 0.7437    Time elaps: 00:00:55    Time remain: 00:15:54.

Epc: 1/1    Bat: 0710/12662    Dis Loss: 1.1867    Gen Loss: 2.5247    Time elaps: 00:00:56    Time remain: 00:15:59.
Epc: 1/1    Bat: 0720/12662    Dis Loss: 1.2205    Gen Loss: 0.4711    Time elaps: 00:00:57    Time remain: 00:15:57.
Epc: 1/1    Bat: 0730/12662    Dis Loss: 0.9957    Gen Loss: 2.3570    Time elaps: 00:00:58    Time remain: 00:15:55.
Epc: 1/1    Bat: 0740/12662    Dis Loss: 0.9696    Gen Loss: 0.7305    Time elaps: 00:00:59    Time remain: 00:15:54.
Epc: 1/1    Bat: 0750/12662    Dis Loss: 0.7118    Gen Loss: 1.3803    Time elaps: 00:00:59    Time remain: 00:15:52.
Epc: 1/1    Bat: 0760/12662    Dis Loss: 0.7068    Gen Loss: 1.1431    Time elaps: 00:01:00    Time remain: 00:15:50.
Epc: 1/1    Bat: 0770/12662    Dis Loss: 0.6253    Gen Loss: 1.8779    Time elaps: 00:01:01    Time remain: 00:15:49.
Epc: 1/1    Bat: 0780/12662    Dis Loss: 0.6859    Gen Loss: 0.9399    Time elaps: 00:01:02    Time remain: 00:15:47.
Epc: 1/1    Bat: 0790/12662    Dis Loss: 0.9305    Gen Loss: 1.0426    Time elaps: 00:01:02    Time remain: 00:15:46.
Epc: 1/1    Bat: 0800/12662    Dis Loss: 1.1372    Gen Loss: 1.0134    Time elaps: 00:01:03    Time remain: 00:15:45.

Epc: 1/1    Bat: 0810/12662    Dis Loss: 1.1003    Gen Loss: 0.5492    Time elaps: 00:01:04    Time remain: 00:15:49.
Epc: 1/1    Bat: 0820/12662    Dis Loss: 0.8950    Gen Loss: 1.2761    Time elaps: 00:01:05    Time remain: 00:15:47.
Epc: 1/1    Bat: 0830/12662    Dis Loss: 0.6230    Gen Loss: 1.4361    Time elaps: 00:01:06    Time remain: 00:15:46.
Epc: 1/1    Bat: 0840/12662    Dis Loss: 0.7917    Gen Loss: 0.8380    Time elaps: 00:01:07    Time remain: 00:15:44.
Epc: 1/1    Bat: 0850/12662    Dis Loss: 0.9547    Gen Loss: 2.2209    Time elaps: 00:01:07    Time remain: 00:15:42.
Epc: 1/1    Bat: 0860/12662    Dis Loss: 0.7468    Gen Loss: 1.0189    Time elaps: 00:01:08    Time remain: 00:15:41.
Epc: 1/1    Bat: 0870/12662    Dis Loss: 0.9362    Gen Loss: 0.8628    Time elaps: 00:01:09    Time remain: 00:15:39.
Epc: 1/1    Bat: 0880/12662    Dis Loss: 1.0218    Gen Loss: 0.6039    Time elaps: 00:01:10    Time remain: 00:15:38.
Epc: 1/1    Bat: 0890/12662    Dis Loss: 0.9555    Gen Loss: 0.7966    Time elaps: 00:01:10    Time remain: 00:15:36.
Epc: 1/1    Bat: 0900/12662    Dis Loss: 0.8484    Gen Loss: 0.9175    Time elaps: 00:01:11    Time remain: 00:15:35.

Epc: 1/1    Bat: 0910/12662    Dis Loss: 0.8950    Gen Loss: 1.1498    Time elaps: 00:01:12    Time remain: 00:15:39.
Epc: 1/1    Bat: 0920/12662    Dis Loss: 0.6742    Gen Loss: 1.1909    Time elaps: 00:01:13    Time remain: 00:15:38.
Epc: 1/1    Bat: 0930/12662    Dis Loss: 1.2154    Gen Loss: 0.6801    Time elaps: 00:01:14    Time remain: 00:15:36.
Epc: 1/1    Bat: 0940/12662    Dis Loss: 0.8548    Gen Loss: 0.8637    Time elaps: 00:01:14    Time remain: 00:15:35.
Epc: 1/1    Bat: 0950/12662    Dis Loss: 0.7354    Gen Loss: 1.8802    Time elaps: 00:01:15    Time remain: 00:15:33.
Epc: 1/1    Bat: 0960/12662    Dis Loss: 0.7869    Gen Loss: 0.8962    Time elaps: 00:01:16    Time remain: 00:15:32.
Epc: 1/1    Bat: 0970/12662    Dis Loss: 0.8865    Gen Loss: 0.7047    Time elaps: 00:01:17    Time remain: 00:15:30.
Epc: 1/1    Bat: 0980/12662    Dis Loss: 0.9146    Gen Loss: 0.8000    Time elaps: 00:01:17    Time remain: 00:15:29.
Epc: 1/1    Bat: 0990/12662    Dis Loss: 0.7477    Gen Loss: 1.6475    Time elaps: 00:01:18    Time remain: 00:15:27.
Epc: 1/1    Bat: 1000/12662    Dis Loss: 1.1619    Gen Loss: 0.5018    Time elaps: 00:01:19    Time remain: 00:15:26.

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.